HDMI (i2s) Cable Recommendation


I think that optical may well work best as a digital interconnect on my DSD in my system. . . but unfortunately the DMP does not have an optical output, and I’d have to use the HDMI for the many SACDs I listen to. I have used the optical for my DVR and for my Blu-ray player and the sound is wonderful with a Lifatec glass optical cable.


Toslink optical is a narrow bandwidth high jitter interconnection the plastic in the jumpers is not as good as real glass used in ST- glass connections.


The DSD seems to deal with any jitter very well, and there’s no plastic in the cable I’m using. Bandwidth is sufficient for my material.


lonson is right. FWIW With a good TOSLink cable (not the cheap things that come with VCRs, DVRs, Blu-ray, etc.) TOSLink can often do 24/192k, not as high as the other inputs, but high enough for many. And yes, I like ST-glass, etc. much better, but it’s not very standard (at least the protocol used by consumer audio land isn’t very standard.) The jitter in TOSLink matters much less to the DS than most DACs, multiple people have been surprised at how good it sounds.


Toslink would only be a last resort for me.


Well that was what I thought years ago, accepting the so-called “conventional wisdom.” But my ears have shown me that with a very good cable Toslink can be superb.


I guess I am too OCD to accept the use of the poor connectors and the cheap optical to electrical and back converters to allow myself to use toslink in my audio system. Is it good enough for TV etc. yes I guess. I have I2S so why would I use any less! SPDIF is in itself a heavy compromise and than to put it through another conversion, no thanks.


Suit yourself. Not all optical cables are cheaply made or have cheap connections however. Just as not all HDMI cables are well-made.



Belden Series FE cables work great for me.


I guess you missed my point. The Toslink connector is very poorly designed not just the ones on the cable but the ones on the inputs and outputs of the devices that are to be connected. The opto converters themselves are of very low quality too so no matter how much you spend on a real glass toslink cable you still end up with a compromised connection. I never said all HDMI cables are well made. The one that comes with the DMP is. besides The difference between SPDIF and I2S is like the difference between composite and colorstream connections on a monitor. One connection mashes all the clocks and data into one stream that must be reconstructed on the receiving end. It is surprising how the Directstream Dac is supposed to be immune to jitter and yet how easy it is to hear the difference between something connected via I2S and toslink or SPDIF/AESEBU. The sound quality improvement of the DMP over the PWT is also an example of how important it is to reduce jitter to vanishing levels. As you say to each his own.


I don’t think I missed your point. Regardless, my ears in my system tell me that a well-made glass optical cable allows a resulting sound that rivals that resulting from the I2S connection in my system in the DSD. I know that the conversion system is said to be inferior but to my ears that doesn’t have to mean inferior sound. I prefer it to the sound with most coaxial cables I’ve tried, and it competes very favorably with I2S via HDMI in my system. To each their own.


Most people that do A/B tests with TOSLink (or any interconnect) don’t disconnect all other interconnects to keep their groundloops. radio emissions, etc. from interfering with the cable under test. In essence you are getting the worse from all of your cables (except for jitter) by not comparing cables in isolation. Conversely you can hook up all possible interconnect from something like the PWT to the DS and when you switch inputs you are only comparing jitter since all of the interconnect then share the same groundloops and RFI environment… The DMP doesn’t have a TOSLink connector, but you can compare RCA, XLR and I2S on it. Often people don’t think you are really changing the inputs in that scenario.

Don’t assume simplistic explanations like multiplexed clocks (or the lack thereof) are really capturing the relevant differences between interconnects especially on DACs (or other equipment) that use different operating principles that your experience has prepared you for. Also don’t assume everyone else’s system will react the same to these kind of experiments.


Perhaps some explanation might help:

The DS doesn’t use the incoming clocks from any input to read or decode that input’s data, this matters in two ways:

There are no PLLs, FLLs, etc that track any input so jitter doesn’t transfer from the inputs to the output clocking. i.e. there’s nothing that looks at edges of the clock or the data signals in the incoming streams.

For the data it oversamples all inputs using the same high speed clock (approx. 170MHz) and then does pattern matching to discover the zeros and ones. By using a fast clock it will get multiple samples in a row for any high or low in the incoming streams: it then looks near the center of a bit to see if it’s high or low. In the case of I2S that’s about it, with S/PDIF, AES and TOSLink it’s looking for wide vs narrow pulses for zeros or ones, still it reads them in the middle. Simple pattern matching.

After the samples are decoded they are put into the expected elastic buffer…

Also FWIW all inputs are always done in parallel no matter which one is currently selected.

Perhaps this helps explain why the DS isn’t so sensitive to jitter as most DACs. I’m not saying no-one will ever hear a jitter difference, but I haven’t encountered anyone personally that can reliably tell me which input is currently selected with, say, the PWT as a source and all inputs connected.

I always use TOSLink for DS demos at someone else’s house since it isolates the DS from more system interactions. (I bring a cheap RCA to TOSLink converter and a cheap USB -> RCA converter so that I can play from most systems with some chain of conversions. A lot of DACs can’t even lock on the signals thru that chain of kludges, but on the DS it doesn’t affect the sound quality enough to try to do something “that makes more sense.”


I bought a cardas and as with all my other cardas cables have enjoyed it.


A couple years into my enjoyment of my DS Sr and DMP, I decided to try a PS Audio AC12 I2S cable after Lonson’s strong recommendation. I’ve been using an Audioquest Coffee. I thought that was noticeably better than the stock cable, even after going back and forth a couple times.
Well, the AC12 is much livelier, more detailed, enjoyable than the Coffee. Although I can’t work out how this can happen, it does make a very nice difference; I guess maybe the Coffee was decaf…


Hi @RonP, where does one find an AC12 I2S cable at this point? Thanks.


You can set up a search on Hifishark.com and it will send you a daily report of any postings. It’s a search aggregator.


I just googled it and found it at a PSA dealer, Jaguar Audio in Seattle. They said it was their demo cable, although it and it’s packaging look new.
I was looking at the Wireworld Platinum but no way would I spring for that price. But for what I paid for the AC12, the improvement gives me good value for money.


Thanks for the tip. I am familiar with HiFiShark. I just checked and the lone North American AC12 I2S listing it aggregated was 3.5 years ago. I guess I will stick with my $12 HDMI cable. It sounds awesome to me and I am blissfully ignorant of anything better out there!


Thanks for the tip. I just looked and it appears to be gone (sold?). Also, three meters is way more than I need. I would be looking for three FEET or thereabouts. Thanks again.